Positron emission tomography (PET) and magnetic resonance imaging (MRI) are two of the most powerful imaging modalities currently in use for neurological studies. Recently, scanners capable of simultaneous PET and MR data acquisition in human subjects have become a reality and this new technology opens up possibilities impossible to realize using sequentially acquired data. One such example is using the MR data for improving the performance of the PET scanner. While PET as a technique has many advantages, including the fact that it could potentially provide a quantitative means to assess in vivo biological processes, the accuracy of the PET measurements is confounded by several factors. For example, attenuation and scatter correction have to be performed to account for the interactions of the gamma-ray photons in the subject before reaching the detectors; motion correction has to be applied to avoid the degradation of the images due to involuntary head movements; partial volume effect correction is required due to the relatively limited spatial resolution; the radiotracer arterial input function is required for kinetic modeling. The spatially and temporally correlated MR data acquired simultaneously offer the unique opportunity to correct for these confounding effects and improve the reliability and reproducibility of the PET estimates. Although many neurological applications could benefit from these methodological improvements, in this proposal we are focusing on Alzheimer's disease (AD) for demonstrating the potential of improved MR-PET quantification. MRI and PET are widely used and provide largely complementary information in assessment of AD patients. Equally important, AD is a great test situation for the development of MR-PET because the confounding factors mentioned above are especially important in this patient population and are a substantial limitation of existing PET research. Specifically, we will: (1) Develop and validate an accurate MR-based head attenuation correction method. We hypothesize that using novel sequences for imaging the bone tissue and improved methods for combining these with high resolution anatomical MR, head attenuation maps more accurate than those obtained from segmented CT can be obtained; (2) Improve the quantification of PET data using the simultaneously acquired MR data. We hypothesize that the temporally and spatially correlated MR data will allow us to improve the reliability of the PET data by performing motion and partial volume effect corrections and estimating the radiotracer arterial input function; (3) Evaluate the added value of MR-optimized PET measurements as biomarkers of disease progression in AD. We hypothesize that the effect size of PET differences would be increased and the variability in PET measurements would be decreased after MR-optimization.